agent-codinglanguage-mapper 1.1.3__tar.gz → 1.1.4__tar.gz

{agent_codinglanguage_mapper-1.1.3/src/agent_codinglanguage_mapper.egg-info → agent_codinglanguage_mapper-1.1.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agent-codinglanguage-mapper
-Version: 1.1.3
+Version: 1.1.4
 Summary: Agent-focused semantic mapping and language services for mixed-language codebases.
 Author: Dark Light
 License-Expression: MPL-2.0
@@ -106,6 +106,12 @@ Parser-native nodes are intentionally private. Ranges are zero-based and half-op
 
 `CodebaseMapper.open()` creates an outline index first. `inspect`, `focus`, and outgoing `callees`/`uses` traces load detail only for the selected source file. Incoming `references`/`callers`/`used_by` traces load the remaining workspace detail because their evidence can originate in any indexed file. Parsed detail and focus remain resident for the mapper session. The same optimized path is used for small and large files.
 
+Batch consumers that only need the complete outline should use
+`snapshot = CodebaseMapper.index(".")`. This one-shot path returns the same deterministic
+`WorkspaceSnapshot` without retaining source buffers, parser outlines, or the interactive refresh
+cache after indexing. It is intended for inventory and documentation pipelines over very large
+repositories.
+
 The private session index retains each parser-native outline tree, its source buffer, a source digest, and a lightweight file fingerprint. A detail upgrade therefore queries the existing tree instead of rereading or reparsing every indexed source. When an indexed source changes during a worker session, the mapper rebuilds the outline snapshot, updates `workspace_revision`, clears stale detail/focus state, and invalidates old cursors before answering the next request. Parser-native nodes never appear in the public result types.
 
 ## CLI

{agent_codinglanguage_mapper-1.1.3 → agent_codinglanguage_mapper-1.1.4}/README.md

@@ -59,6 +59,12 @@ Parser-native nodes are intentionally private. Ranges are zero-based and half-op
 
 `CodebaseMapper.open()` creates an outline index first. `inspect`, `focus`, and outgoing `callees`/`uses` traces load detail only for the selected source file. Incoming `references`/`callers`/`used_by` traces load the remaining workspace detail because their evidence can originate in any indexed file. Parsed detail and focus remain resident for the mapper session. The same optimized path is used for small and large files.
 
+Batch consumers that only need the complete outline should use
+`snapshot = CodebaseMapper.index(".")`. This one-shot path returns the same deterministic
+`WorkspaceSnapshot` without retaining source buffers, parser outlines, or the interactive refresh
+cache after indexing. It is intended for inventory and documentation pipelines over very large
+repositories.
+
 The private session index retains each parser-native outline tree, its source buffer, a source digest, and a lightweight file fingerprint. A detail upgrade therefore queries the existing tree instead of rereading or reparsing every indexed source. When an indexed source changes during a worker session, the mapper rebuilds the outline snapshot, updates `workspace_revision`, clears stale detail/focus state, and invalidates old cursors before answering the next request. Parser-native nodes never appear in the public result types.
 
 ## CLI

{agent_codinglanguage_mapper-1.1.3 → agent_codinglanguage_mapper-1.1.4}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "agent-codinglanguage-mapper"
-version = "1.1.3"
+version = "1.1.4"
 description = "Agent-focused semantic mapping and language services for mixed-language codebases."
 readme = "README.md"
 requires-python = ">=3.10,<3.15"

agent_codinglanguage_mapper-1.1.4/src/agent_codinglanguage_mapper/_version.py

@@ -0,0 +1 @@
+__version__ = "1.1.4"

{agent_codinglanguage_mapper-1.1.3 → agent_codinglanguage_mapper-1.1.4}/src/agent_codinglanguage_mapper/indexer.py

@@ -81,7 +81,7 @@ class WorkspaceIndexCache:
 
 
 def _deduplicate_parser_symbols(items: tuple[Symbol, ...]) -> tuple[Symbol, ...]:
-    """Collapse C/C++ typedef aliases emitted for the same concrete declaration."""
+    """Collapse identical symbols and C/C++ aliases for one concrete declaration."""
 
     def declaration_key(item: Symbol) -> tuple[str, str, int, int]:
         anchor = item.name_range or item.range
@@ -103,7 +103,35 @@ def _deduplicate_parser_symbols(items: tuple[Symbol, ...]) -> tuple[Symbol, ...]
         if key in emitted:
             continue
         emitted.add(key)
-        group = groups[key]
+        group: list[Symbol] = []
+        identities: set[tuple[object, ...]] = set()
+        for candidate in groups[key]:
+            identity = (
+                candidate.name,
+                candidate.qualified_name,
+                candidate.kind,
+                candidate.language,
+                candidate.range,
+                candidate.project_id,
+                candidate.file_id,
+                candidate.signature,
+                candidate.native_kind,
+                candidate.parent_id,
+                candidate.name_range,
+                candidate.role,
+            )
+            if identity in identities:
+                continue
+            identities.add(identity)
+            group.append(candidate)
+        if group and group[0].language is Language.DELPHI:
+            logical: dict[tuple[str, str, str], Symbol] = {}
+            for candidate in group:
+                logical.setdefault(
+                    (candidate.kind, candidate.role, candidate.parent_id),
+                    candidate,
+                )
+            group = list(logical.values())
         concrete = [candidate for candidate in group if candidate.kind in concrete_kinds]
         aliases = [candidate for candidate in group if candidate.kind == "type_alias"]
         if len(concrete) == 1 and aliases:
@@ -631,6 +659,7 @@ def _workspace_symbol_id(project_id: str, file_id: str, symbol: Symbol) -> str:
         "symbol",
         project_id,
         file_id,
+        anchor.path,
         symbol.qualified_name,
         str(anchor.start_line),
         str(anchor.start_character),

{agent_codinglanguage_mapper-1.1.3 → agent_codinglanguage_mapper-1.1.4}/src/agent_codinglanguage_mapper/mapper.py

@@ -141,7 +141,11 @@ class CodebaseMapper:
 
     @classmethod
     def open(cls, root: str | Path, languages: str | tuple[Language, ...] = "auto") -> "CodebaseMapper":
-        selected = None if languages == "auto" else tuple(Language(item) if isinstance(item, str) else item for item in languages)
+        selected = (
+            None
+            if languages == "auto"
+            else tuple(Language(item) if isinstance(item, str) else item for item in languages)
+        )
         index_cache = WorkspaceIndexCache()
         return cls(
             root,
@@ -151,6 +155,26 @@ class CodebaseMapper:
             index_cache=index_cache,
         )
 
+    @classmethod
+    def index(
+        cls,
+        root: str | Path,
+        languages: str | tuple[Language, ...] = "auto",
+    ) -> WorkspaceSnapshot:
+        """Build a complete outline snapshot without retaining an interactive cache."""
+
+        selected = (
+            None
+            if languages == "auto"
+            else tuple(Language(item) if isinstance(item, str) else item for item in languages)
+        )
+        return build_workspace_snapshot(
+            root,
+            selected,
+            detail_files=(),
+            index_cache=None,
+        )
+
     def __enter__(self) -> "CodebaseMapper":
         return self
 
@@ -189,10 +213,7 @@ class CodebaseMapper:
                 if (
                     request.file_id
                     and candidate.file_id != request.file_id
-                    and (
-                        requested_source is None
-                        or Path(candidate.range.path).as_posix() != requested_source.path
-                    )
+                    and (requested_source is None or Path(candidate.range.path).as_posix() != requested_source.path)
                 ):
                     continue
                 if request.languages and candidate.language not in request.languages:
@@ -273,11 +294,7 @@ class CodebaseMapper:
         )
 
     def _refresh_changed_sources(self) -> None:
-        if (
-            not self._lazy_details
-            or self._index_cache is None
-            or not self._index_cache.sources_changed(self.root)
-        ):
+        if not self._lazy_details or self._index_cache is None or not self._index_cache.sources_changed(self.root):
             return
         self.snapshot = build_workspace_snapshot(
             self.root,
@@ -351,7 +368,7 @@ class CodebaseMapper:
             text_offset = self._decode_cursor(request)
             if text_offset > len(full_text) or (request.cursor and text_offset == len(full_text)):
                 raise ProtocolError("cursor offset is invalid")
-            item[text_key] = full_text[text_offset:text_offset + _SOURCE_CHUNK_CHARS]
+            item[text_key] = full_text[text_offset : text_offset + _SOURCE_CHUNK_CHARS]
             if text_key == "source":
                 line_offset = full_text[:text_offset].count("\n")
                 item["citation"] = f"{symbol.range.path}:{int(str(item['citation']).rsplit(':', 1)[1]) + line_offset}"
@@ -450,7 +467,7 @@ class CodebaseMapper:
                     low = middle + 1
                 else:
                     high = middle - 1
-            bounded[detail_key] = text[:max(0, high)]
+            bounded[detail_key] = text[: max(0, high)]
             return bounded
 
         for key in ("kind", "role", "project_id", "file_id", "qualified_name", "signature"):
@@ -487,9 +504,7 @@ class CodebaseMapper:
             selected = [item for item in references if item.target_id in request_target_id_set]
             return self._page_response(request, [_reference_mapping(item) for item in selected])
         if relation in {"callers", "used_by"}:
-            target_ids = {
-                item.source_id for item in references if item.target_id in request_target_id_set
-            }
+            target_ids = {item.source_id for item in references if item.target_id in request_target_id_set}
             target_ids.update(
                 item.source_id
                 for item in self.snapshot.relations
@@ -501,16 +516,12 @@ class CodebaseMapper:
             return self._page_response(request, [_symbol_mapping(item) for item in caller_symbols])
         if relation in {"callees", "uses"}:
             target_ids = {
-                item.target_id
-                for item in references
-                if item.source_id in request_target_id_set and item.target_id
+                item.target_id for item in references if item.source_id in request_target_id_set and item.target_id
             }
             target_ids.update(
                 item.target_id
                 for item in self.snapshot.relations
-                if relation == "uses"
-                and item.source_id in request_target_id_set
-                and item.target_id
+                if relation == "uses" and item.source_id in request_target_id_set and item.target_id
             )
             callee_symbols = [
                 self.snapshot.symbols_by_id[item] for item in target_ids if item in self.snapshot.symbols_by_id
@@ -545,7 +556,7 @@ class CodebaseMapper:
         start = self._decode_cursor(request)
         selected: list[dict[str, object]] = []
         used = 0
-        for item in items[start:start + request.max_items]:
+        for item in items[start : start + request.max_items]:
             size = len(json.dumps(item, sort_keys=True, separators=(",", ":")))
             if used + size > request.max_chars:
                 if not selected:

{agent_codinglanguage_mapper-1.1.3 → agent_codinglanguage_mapper-1.1.4/src/agent_codinglanguage_mapper.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agent-codinglanguage-mapper
-Version: 1.1.3
+Version: 1.1.4
 Summary: Agent-focused semantic mapping and language services for mixed-language codebases.
 Author: Dark Light
 License-Expression: MPL-2.0
@@ -106,6 +106,12 @@ Parser-native nodes are intentionally private. Ranges are zero-based and half-op
 
 `CodebaseMapper.open()` creates an outline index first. `inspect`, `focus`, and outgoing `callees`/`uses` traces load detail only for the selected source file. Incoming `references`/`callers`/`used_by` traces load the remaining workspace detail because their evidence can originate in any indexed file. Parsed detail and focus remain resident for the mapper session. The same optimized path is used for small and large files.
 
+Batch consumers that only need the complete outline should use
+`snapshot = CodebaseMapper.index(".")`. This one-shot path returns the same deterministic
+`WorkspaceSnapshot` without retaining source buffers, parser outlines, or the interactive refresh
+cache after indexing. It is intended for inventory and documentation pipelines over very large
+repositories.
+
 The private session index retains each parser-native outline tree, its source buffer, a source digest, and a lightweight file fingerprint. A detail upgrade therefore queries the existing tree instead of rereading or reparsing every indexed source. When an indexed source changes during a worker session, the mapper rebuilds the outline snapshot, updates `workspace_revision`, clears stale detail/focus state, and invalidates old cursors before answering the next request. Parser-native nodes never appear in the public result types.
 
 ## CLI

{agent_codinglanguage_mapper-1.1.3 → agent_codinglanguage_mapper-1.1.4}/tests/test_indexer.py

@@ -223,6 +223,40 @@ def test_mapper_open_indexes_languages_and_find_searches_them(tmp_path: Path) ->
     assert {item["citation"] for item in response.result["items"]} == {"modern.py:1", "modern.cs:1"}
 
 
+def test_mapper_index_builds_one_shot_outline_without_workspace_cache(
+    tmp_path: Path,
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    (tmp_path / "service.py").write_text("class Service:\n    pass\n")
+    real_build = indexer_module.build_workspace_snapshot
+    calls: list[tuple[tuple[Language, ...] | None, tuple[str, ...], object | None]] = []
+
+    def recorded_build(
+        root: str | Path,
+        languages: tuple[Language, ...] | None = None,
+        *,
+        detail_files: tuple[str, ...],
+        index_cache: object | None = None,
+    ):
+        calls.append((languages, detail_files, index_cache))
+        return real_build(
+            root,
+            languages,
+            detail_files=detail_files,
+            index_cache=index_cache,
+        )
+
+    monkeypatch.setattr(
+        "agent_codinglanguage_mapper.mapper.build_workspace_snapshot",
+        recorded_build,
+    )
+
+    snapshot = CodebaseMapper.index(tmp_path, languages=(Language.PYTHON,))
+
+    assert any(item.name == "Service" for item in snapshot.symbols)
+    assert calls == [((Language.PYTHON,), (), None)]
+
+
 def test_mapper_refreshes_snapshot_when_source_changes_after_open(tmp_path: Path) -> None:
     source = tmp_path / "changing.py"
     source.write_text("def old_name():\n    return 'old'\n")
@@ -724,6 +758,31 @@ def test_delphi_include_symbol_is_indexed_under_include_file(tmp_path: Path) ->
     assert detail.result["item"]["declaration"] == "procedure IncludedProcedure"
 
 
+def test_delphi_distinct_include_fragments_have_unique_symbol_ids(tmp_path: Path) -> None:
+    (tmp_path / "left.inc").write_text("procedure Configure;\n")
+    (tmp_path / "right.inc").write_text("procedure Configure;\n")
+    (tmp_path / "Host.pas").write_text("unit Host;\ninterface\n{$I left.inc}\n{$I right.inc}\nimplementation\nend.\n")
+
+    snapshot = CodebaseMapper.index(tmp_path)
+    included = [item for item in snapshot.symbols if item.name == "Configure"]
+
+    assert {item.range.path for item in included} == {"left.inc", "right.inc"}
+    assert len({item.target_id for item in included}) == len(included)
+
+
+def test_delphi_repeated_identical_include_symbol_is_indexed_once(tmp_path: Path) -> None:
+    (tmp_path / "shared.inc").write_text("procedure Configure;\n")
+    (tmp_path / "Host.pas").write_text(
+        "unit Host;\ninterface\n{$I shared.inc}\n{$I shared.inc}\nimplementation\nend.\n"
+    )
+
+    snapshot = CodebaseMapper.index(tmp_path)
+    included = [item for item in snapshot.symbols if item.name == "Configure"]
+
+    assert len(included) == 1
+    assert len({item.target_id for item in snapshot.symbols}) == len(snapshot.symbols)
+
+
 def test_delphi_include_change_invalidates_host_outline(tmp_path: Path) -> None:
     include_path = tmp_path / "declarations.inc"
     include_path.write_text("procedure OldIncludedProcedure;\n")

{agent_codinglanguage_mapper-1.1.3 → agent_codinglanguage_mapper-1.1.4}/tests/test_package_contract.py

@@ -53,7 +53,7 @@ def test_release_metadata_is_stable_and_cross_platform() -> None:
         project = tomllib.load(stream)["project"]
 
     assert project["name"] == "agent-codinglanguage-mapper"
-    assert project["version"] == "1.1.3"
+    assert project["version"] == "1.1.4"
     assert project["authors"] == [{"name": "Dark Light"}]
     assert project["requires-python"] == ">=3.10,<3.15"
     assert "Operating System :: MacOS" in project["classifiers"]
@@ -76,7 +76,7 @@ def test_local_release_gates_include_guideline_tests() -> None:
     with (ROOT / "pyproject.toml").open("rb") as stream:
         configuration = tomllib.load(stream)
 
-    assert configuration["project"]["version"] == "1.1.3"
+    assert configuration["project"]["version"] == "1.1.4"
     assert (ROOT / "tests/test_guideline_formatting.py").is_file()
     assert (ROOT / "tests/test_guidelines.py").is_file()
 
@@ -102,7 +102,7 @@ def test_public_api_uses_only_generic_names() -> None:
         __version__,
     )
 
-    assert __version__ == "1.1.3"
+    assert __version__ == "1.1.4"
     for exported in (
         AgentRequest,
         AgentResponse,

agent_codinglanguage_mapper-1.1.3/src/agent_codinglanguage_mapper/_version.py

@@ -1 +0,0 @@
-__version__ = "1.1.3"